[gnu.g++.bug] spurious warning about MAXDOUBLE

kenny@CS.UIUC.EDU ("Kevin B. Kenny [the Arch-Traitor] KE9TV") (01/24/90)

System: Sun 3/50 (software floating point (-fsoft-float by default))
Compiler: G++ 1.36.1

The following program generates a message about MINDOUBLE being out of
range.  It seems to work at runtime.

% cat -n mindub.cc
     1	#include <stream.h>
     2	#include <values.h>
     3	
     4	main (int argc, char **argv) {
     5	  cout << MINDOUBLE << "\n";
     6	}
% g++ -o mindub -v -Wall -g -O mindub.cc
gcc version 1.36.1 (based on GCC 1.36)
 /home/klin/kenny/lib/gcc-cpp -+ -v -undef -D__GNUC__ -D__GNUG__ -D__cplusplus -Dmc68000 -Dsun -Dunix -D__mc68000__ -D__sun__ -D__unix__ -D__OPTIMIZE__ -Wall -Dmc68020 mindub.cc /usr/tmp/cca18444.cpp
GNU CPP version 1.36
 /home/klin/kenny/lib/gcc-cc1plus /usr/tmp/cca18444.cpp -quiet -dumpbase mindub.cc -g -O -Wall -version -o /usr/tmp/cca18444.s
GNU C++ version 1.36.1 (based on GCC 1.36) (68k, MIT syntax) compiled by GNU C version 1.36.
default target switches: -m68020 -mc68020 -mbitfield
mindub.cc: In function int main (int, char **):
mindub.cc:5: warning: floating point number exceeds range of `double'
 /home/klin/kenny/lib/gcc-as -mc68020 -o mindub.o /usr/tmp/cca18444.s
 /home/klin/kenny/lib/gcc-ld -o mindub -e start -dc -dp -Bstatic /lib/crt0.o /lib/Fcrt1.o mindub.o -lg++ /home/klin/kenny/lib/gcc-gnulib -lg -lc
% mindub
4.94066e-324     <- looks correct...
% 

Kevin, KE9TV